1

DPWA Winners

2019

2

TABLE OF CONTENTS SABER M. ELSAYED: ‘New Integral Modelling and Analysis Approach for Storm Surge-Induced Barrier Breaching, Coastal Inundation and Subsequent Vertical Saltwater Intrusion.’ 1 INTRODUCTION ............................................................................................................................. 3 2 THE XBEACH-SEAWAT MODEL SYSTEM ................................................................................... 4 3 BREACHING OF CBS: LIMITATIONS AND PROPOSED NOVEL PROCESSES ......................... 5 4 COMBINED MODELLING OF COASTAL BARRIER BREACHING AND INUNDATION ............... 8 5 IMPLICATIONS OF COASTAL FLOODS FOR GROUNDWATER CONTAMINATION: THE CASE OF NEAR BREMERHAVEN ....................................................................................... 11 6 SUBSURFACE DRAINAGE EFFECT ON THE RESILIENCE OF COASTAL AQUIFERS .......... 14 7 SUMMARY, LESSONS LEARNT AND IMPLICATIONS FOR PRACTICAL APPLICATIONS ..... 16 8 ACKNOWLEDGEMENT ................................................................................................................ 17 9 REFERENCES .............................................................................................................................. 17

Summary ............................................................................................................................................... 18 Résumé ................................................................................................................................................. 18 Zusammenfassung ................................................................................................................................ 18 Resumen ............................................................................................................................................... 20

YOSHINOSUKE KURAHARA: 'Prediction of Shackle Motion Hanged from a Jib Top of Crane Barge by a Coupling Numerical Model of Three Motions'

1 INTRODUCTION ........................................................................................................................... 21 2 NUMERICAL MODEL .................................................................................................................... 22 3 HYDRAULIC EXPERIMENTS ....................................................................................................... 28 4 DISCUSSION ................................................................................................................................ 31 5 CONCLUSIONS ............................................................................................................................ 34 6 ACKNOWLEDGMENTS ................................................................................................................ 34 7 REFERENCES .............................................................................................................................. 34

Summary ............................................................................................................................................... 35 Résumé ................................................................................................................................................. 35 Zusammenfassung ................................................................................................................................ 36 Resumen ............................................................................................................................................... 36 Javier Murgoitio Esandi: 'Assessment of Overtopping Vertical River Walls due to Vessel-Generated

Waves (Vessel Wash)'

1 INTRODUCTION ........................................................................................................................... 37 2 BACKGROUND ............................................................................................................................. 37 3 PROPOSED METHODOLOGICAL APPROACH .......................................................................... 40 4 SUMMARY .................................................................................................................................... 47 5 SYMBOLS ..................................................................................................................................... 47 6 REFERENCES .............................................................................................................................. 47

Summary ............................................................................................................................................... 49 Résumé ................................................................................................................................................. 49 Zusammenfassung ................................................................................................................................ 49 Resumen ............................................................................................................................................... 50

3

PIANC De Paepe-Willems Award Winner 2019

NEW INTEGRAL MODELLING AND ANALYSIS APPROACH FOR STORM

SURGE-INDUCED BARRIER BREACHING, COASTAL INUNDATION AND

SUBSEQUENT VERTICAL SALTWATER INTRUSION

SABER M. ELSAYED Division of Hydromechanics and Coastal Engineering, Leichtweiß Institute for Hydraulic Engineering and Water Resources, TU-Braunschweig, Braunschweig, Germany, Email: s-m.elsayed@tu-braunschweig.de

Keywords: Coastal barriers breaching; Coastal inundation; Storm-driven saltwater intrusion; Subsurface drainage. Mots-clés : Brèche des barrières côtières ; Inondation côtière ; Intrusion d'eau salée due aux tempêtes ; Drainage souterrain.

1 INTRODUCTION

Natural disasters (e.g. extreme storm surges) and consequences of climate change on coastal areas represent serious threats to safety of coastal defences as well as to coastal groundwater resources. Due to global warming and possible increase of frequency and intensity of coastal storms, many coastal systems may experience accelerated erosion, barrier breaching, flooding and subsequent VSWI into coastal aquifers. Natural CBS as indicated in Fig (1.a) represent an important component of the defence system against such threats and their possible induced catastrophic consequences to human, strategic infrastructure, coastal ecosystems and landscapes. However, during extreme surges, water levels may increase from moderate sea level (MSL) as schematically shown in Fig (1.b) under the effect of wind speed and low pressure to higher surge level. As a result, barriers become directly attacked by shortwaves riding on the surge, leading to barrier erosion. Subsequently, barriers may breach, inducing coastal flooding and subsequent VSWI.

(a)

(b)

Fig. 1: Surface and subsurface processes at sea/land boundary: (a) during moderate sea conditions and (b) during extreme surges

4

Though breaching/overtopping, subsequent flooding and VSWI are naturally successive and hydraulically interconnected processes, the complexity of these processes has led to split their modelling. As a result, no integral model yet exists to reliably assess the vulnerability of coastal defences to breaching under extreme storm surges and to draw the implications of breaching-induced coastal floods for the contamination of coastal aquifers, which represent important water resources. Coastal aquifers are extremely sensitive to the VSWI and their contaminations may last for decades until they are restored naturally [Elsayed and Oumeraci, 2018]. Indeed, the regional flow induced by difference in hydraulic gradients (Fig 1) supports such a long-term aquifers restoration/remediation. Nevertheless, sustainable development of coastal areas might also be affected owing to the long-lasting remediation and the need for more treatment costs. Therefore, there is an urgent need to an integral modelling/prediction tool that can successively and reliably (i) simulate the breaching/overtopping of earthen CBS, (ii) predict the inland discharge and its propagation in the hinterland and (iii) accurately estimate extent of VSWI and related aquifers remediation. Thus, an appropriate mitigation measure might be suggested to mitigate the VSWI and to shorten the natural remediation time. Therefore, the main objectives of this study are (i) to provide an improved understanding of breaching of CBS, induced inundation and subsequent VSWI, (ii) to address the modelling of these processes in an integral and well-validated approach, (iii) to draw the implications of coastal floods for groundwater contamination, and (iv) to examine the suitability and performance of an SSDN as a mitigation measure of the VSWI. To achieve these objectives, a new modelling approach [Elsayed, 2017] is developed. The proposed approach utilises an improved version of the widely used open-source code XBeach [Roelvink et al., 2009] for simulating overtopping/breaching of CBS and subsequent flooding and SEAWAT [Langevin et al., 2008] for simulating the VSWI. The components of the model system are systematically validated with large-scale tests for dune erosion and further diverse data from the literature, and the entire approach is applied to a case study. The rationale behind this model system is discussed below.

2 THE XBEACH-SEAWAT MODEL SYSTEM

Breaching of CBS represents an important source of coastal flooding, where breach-induced inlets work as pathways to inland inundation and subsequent VSWI. Breaching is a complex hydro-geo- morphodynamic process, commonly initiated when water overflows a depressed portion in a protective barrier. Given sufficient duration and intensity, the flow will induce an inlet that causes flow across the barrier. During storm surges, the processes that may initiate a breach (Fig. 2) are: (i) Impact of breaking waves, (ii) Wave run-up and rundown, (iii) wave overtopping/overwash, (iv) Overflow of combined waves and surge and (v) seepage and piping. The first two processes may initiate a breach from the seaward while the rest may initiate a breach from the landward.

Fig. 2: Hydrodynamic processes that may initiate a coastal barrier breach

5

As a result, a breach may widen due to the hydrodynamic interaction with sediments and the induced sediment transport from the breach bed and wedges and the subsequent soil avalanching due to slope instabilities. Therefore, rather than starting with developing a new breaching model to simulate these diverse processes, selection of a state-of-the- art breaching model is favoured. As a result, XBeach is selected as the most appropriate model to simulate these processes.

XBeach has sufficiently proved his capability to predict overtopping rates as compared to empirical overtopping models (e.g. EurOtop), which is also unsuitable to calculate inland discharges through breaches because of the dynamic nature of the breaching process [Elsayed and Oumeraci, 2017]. Moreover, XBeach is capable, as demonstrated in Elsayed and Oumeraci (2016), to simulate the breaching process and induced inundation in a single simulation of the coastal zone and the adjacent hinterland so that mutual interactions between both processes are also considered (See Section 4).

Thus, XBeach represents currently the most appropriate tool to predict breaching dimensions and inland discharges required to reliably predict extents of inundation and VSWI. Nevertheless, due to its simple groundwater module, XBeach cannot account for subsurface mass transport. Therefore, XBeach is unidirectionally coupled with SEAWAT [Langevin et al., 2008] as a subsurface model, which is capable to simulate fully coupled groundwater flow and solute transport in porous media. Thus, the XBeach- SEAWAT approach represents an integral surface-subsurface model system capable of (i) simulating coastal erosion and breaching as well as the induced inundation and (ii) considering the implications of breaching/overtopping-induced inundation for groundwater contamination and subsequent natural remediation. SEAWAT is selected among other solute transport models (e.g. SUTRA) because it includes a ready to use drainage package that can be adapted to simulate the subsurface drainage as a measure to mitigate the VSWI as elaborated in the following sections.

3 BREACHING OF CBS: LIMITATIONS AND PROPOSED NOVEL

PROCESSES

3.1 Assessment of XBeach

The performance of XBeach is examined using a unique dataset from 17 large-scale tests for dune erosion (hereafter called GWK-tests), which were performed in the large-scale flume (GWK) in Hannover to physically simulate the erosion of the dunes western of Wangerooge Island, northern Germany. GWK- tests were performed on five cross-shore profiles subject to the same wave conditions (significant height of 1.1 m and peak period of 6.6 s) and different still water levels so that a wide range of wave overtopping rates could be achieved. For all profiles, the dune extension with a frontal slope of 1:1 is built behind the revetment with different dune offset (Fig 3).

(a) (b) (c)

Fig. 3: Physical model setup and dune offset in GWK-tests: (a) without offset, (b) with 3.33 m wide offset and (c) with 6.66 m wide offset

Using a non-modified XBeach (revision 4812), GWK-tests are reproduced. The results (Fig 4) showed a relatively reasonable prediction capability for the scour behind the revetment and the frontal dune erosion, thus illustrating XBeach suitability to simulate coastal erosion. However, the results revealed that the prediction performance got worse for higher overtopping rates (e.g. Fig 4.b for a rate of 423 l/s/m) than for lower wave overtopping rates (e.g. Fig 4.a for a rate of 16.35 l/s/m). This overestimation is verified (Fig 4.c) by comparing the observed and the modelled crest recession for all the 17 GWK-tests.

6

Both crest recession (Fig 4.c) and scour size behind the revetment crest (e.g. Fig 4.a and 4.b) are overestimated by about 70 % in average. However, the mismatch between observed and calculated profiles vanishes with lower overtopping rates and also with increasing the dune offset. Overcoming/reducing such overestimation represent the primary motivation for implementing further physically-based improvements in XBeach before using it in breaching simulations.

3.2 Improvement of XBeach

Rather than using non-physically based limiters to overcome the erosion overestimation, which is common with XBeach for high flow velocity regimes [McCall et al., 2010], and in order to improve its prediction capability in terms of erosion and overwash, two physical processes responsible for these overestimations are identified: (i) the wave nonlinearity effect on sediment transport, which is described in XBeach by a calibration factor for the time-averaged flow due to wave skewness and asymmetry and (ii) the considerable excess of the shear stress actually required to initiate the sediment particle motion as compared to that predicted by the common Shields’ criterion for incipient motion. The importance of examining these two processes arises because they are the governing processes of sediment stirring in XBeach. Therefore, the following related improvements are introduced and implemented in XBeach.

(a) Bed slope effect on sediment transport

The surfbeat mode of XBeach, which is commonly used in practical morphology studies, does not directly simulate the wave shape. Thus, intra-wave processes (e.g. skewness and asymmetry) are implicitly computed. The wave skewness and asymmetry increase the onshore sediment transport and therefore may be one of the reasons behind overestimation of sediment erosion. In order to indirectly account for wave shape within XBeach, a skewness and asymmetry model is used, taking the following form

𝑢𝑎 = 𝛾𝑢𝑎𝑢𝑟𝑚𝑠(𝑆𝑘 − 𝐴𝑠) (1)

Where 𝑢𝑎 is a net velocity that transports sediment onshore under the effect of nonlinear waves; the

skewness 𝑆𝑘 and asymmetry 𝐴𝑠 are parameterised as a function of the Ursell number; 𝑢𝑟𝑚𝑠 is the near-

bed root-mean-squared orbital velocity. The parameter 𝛾𝑢𝑎 is defined in XBeach by the keyword facua and has a default value of 0.1. It represents one of the most important parameters in XBeach as it affects the net cross-shore sediment transport. Based on date collected from the literature (Fig 5.a), it is shown that 𝛾𝑢𝑎 depends on beach slope steepness 𝑆𝑠. As a result, Eq (1) is modified in XBeach as

( 2 )

Eq (2) provides a proper value for 𝑢𝑎, which affects sediment transport rates and thus may contribute to

solve the overestimation problem through stirring more sediment onshore. Therefore, this model improvement proves that onshore sediment transport depends on beach slope.

7

(b) Grain-stabilisation effect on inception of sediment motion

A large part of the erosion overestimation may be attributed to the underestimation of the critical Shields parameter for incipient sediment motion due to the omission of grain-stabilising effects on critical bed shear stresses. Elsayed and Oumeraci (2017) have demonstrated that there are unaccounted forces in the formulation of the critical Shields parameter such as (i) the uprooting force to overcome the sediment interlocking, especially for consolidated/compacted soils and (ii) biological and/or salty stabilisations. The additional shear stresses required to account for these effects are implemented in

XBeach by amplifying the critical Shields parameter 𝜃𝑐 (as schematically shown in Fig 5.b) and hence

the critical stirring velocity 𝑈𝑐𝑟 , using an amplification factor 𝛾𝑝𝑖 as follows

Where 𝐶𝑓 is roughness coefficient and 𝑔 is the gravitational acceleration. 𝑈𝑐𝑟 is mainly a function of the

sediment properties (e.g. mean diameter 𝐷50 and specific gravity s). 𝛾𝑝𝑖 is a new calibration factor for the

grain-stabilisation that considers the increase of the shear stresses required to initiate the sediment motion by amplifying 𝑈𝑐𝑟 to 𝑈𝑐𝑟𝑝𝑖 . Increasing 𝛾𝑝𝑖 increases 𝑈𝑐𝑟𝑝𝑖 and reduces sediment stirring because of

decreasing the mismatch between actual mobilizing flow velocity and the amplified critical velocity

(c)

Fig. 4: Measured and predicted cross-shore profiles of

GWK-tests: (a) with a lower wave overtopping rate of

16.35 l/(s∙m), (b) with a higher wave overtopping rate of

423 l/(s∙m), and (c) modelled vs observed crest

recessions for the 17 GWK-tests

(b) (a)

8

𝑈𝑐𝑟𝑝𝑖 . Through this model modification, it was capable to explain, for the first time, the role of bed soil

stabilisation in sediment stirring and to substitute the use of artificial sediment transport limiters which cannot be physically explained. It indeed justifies why some soils undergo more erosion than others,

based on their grain stabilisation status. Higher values for 𝛾𝑝𝑖 indicate indeed higher soil resistance to

erosion due to grain-stabilisation.

(a)

(b)

Fig. 5: Proposed XBeach improvements: (a) Relation between beach slope steepness and facua and (b) Amplified critical shields parameter to account for the grain-stabilisation

(c) Validation of model modifications

The performance of the improved XBeach was examined for three cases for dune erosion and breaching [Elsayed and Oumeraci, 2017]; among them are the large-scale GWK-tests mentioned above. Meanwhile, since their publication in Elsayed and Oumeraci (2017), the improvements are applied in some studies [Park et al., 2018] to improve the prediction capability of XBeach.

4 COMBINED MODELLING OF COASTAL BARRIER BREACHING AND

INUNDATION

The state-of-the-art modelling of a barrier breaching and the induced inundation is often based on decoupled modelling of these two processes, based on transferring the inland discharge (i.e. inland hydrograph Q(t)) from a breaching/overtopping model to another flood propagation model. Therefore, these traditional decoupled approaches often omit the momentum transfer between both models, leading to incorrect predictions of the flood propagation in the hinterland. In order to consider both mass (flow) and momentum transfer, Elsayed and Oumeraci (2016) demonstrated that XBeach can be applied to simulate the breaching process and the induced inundation in a single model so that mutual interactions and transfer of both mass and momentum are considered. For instance, Fig (6) shows a one-dimensional example for both decoupled (using XBeach as a breaching/overwash model and the famous HEC-RAS model by US Army corps of Engineers as a propagation model) and XBeach alone for combined modelling of breaching/overwash and induced inundation. The inland discharge Q(t) is computed from XBeach at point P2, where Q(t) serves as aninflow boundary condition for the HEC-RAS inundation model for the hinterland. The outcomes of the inundation from XBeach and HEC-RAS are compared at the reference points P3, P4 and P5.

9

Fig. 6: 1-D synthetic cross-shore profile used (a) in XBeach to simulate a 1D overwash event and the induced flood propagation (b) in HEC-RAS where the inland discharge Q(t) is computed from XBeach at point P2 and

used as an inflow boundary condition for flood propagation in the hinterland

Coastal erosion and overwash are forced for the synthetic cross-shore profile in Fig (6) for one hour by synthetic surge and wave conditions. For the decoupled approach, the inland hydrograph is computed at point P2 and transferred to HEC-RAS as an upstream boundary. For the combined modelling by XBeach, the boundaries are ‘automatically’ transferred between both processes. In Fig (7), the water profiles at different times are plotted and water depths and velocities at the points (P3-P5) are computed, showing that the decoupled modelling provides meaningless water levels (water level in the hinterland exceeded the seawater level), especially at times 45 and 60 min due to the omission of the momentum transfer. In fact, flow velocity u(t), which is also crucial as it provides together with Q(t) the momentum, cannot be accounted for in the inflow conditions of the flood propagation model in decoupled approaches. Moreover, the decoupled approach cannot account for the evolution of the inflow width at the upstream boundary in the common inundation models, which also affect the flood kinematics in the hinterland. Such evolution of inflow width arises from the dynamic nature of the breaching process due to widening and deepening with time.

As shown in Fig (7.b and 7.c), the omission of the momentum transfer in the decoupled approach leads to higher calculated water depths at Points (P3-P5) and hence lower flow velocities at the same points. This generally means that decoupled approaches provide incorrect flood extents, water depths and flow kinematics. For this reason, XBeach is suggested as a model for combined modelling of overtopping/ breaching and induced inundation.

In order to demonstrate that XBeach can be applied as a flood propagation model, besides being a model for nearshore hydro-morphodynamics, the mathematical formulations of XBeach and common flood propagation models (e.g. River-2D by University of Alberta and BASEMENT by ETH Zürich) are compared, showing that both models are based on the nonlinear shallow water equations. However,

10

XBeach utilises the Generalised Lagrangian Mean (GLM) approach which applies a more generic representation of the bed shear stresses and the depth-averaged velocities rather than the Eulerian representation that is commonly used in the common propagation models. The GLM approach unambiguously splits a motion into a mean part (Eulerian) and an oscillatory part (Lagrangian), providing a mixed Eulerian-Lagrangian description for the flow field but appointed to fixed Eulerian coordinates. Therefore, GLM applies to any problem, whose governing equations are given in Eulerian form (e.g. common propagation models), with a more thorough representation of the real processes. The latter means that XBeach can also function as a flood propagation model and may, therefore, be applied to simulate, in combination and successively, the breaching/overtopping and induced inundation over a single mesh containing nearshore bathymetry and hinterland topography. As a result, the aforementioned drawbacks of the decoupled approach are overcome.

Fig. 7: Comparison of outcomes using decoupled and combined overwash and inundation modelling

In order to validate the previously improved XBeach for the combined modelling, it is applied to the Het Zwin breaching and inundation test [Elsayed and Oumeraci, 2016]. The outcomes well-illustrated that XBeach is capable to accurately predict breach dimensions and the flow kinematics and depths in the hinterland. The main outcome of this phase is that a well-validated and improved XBeach is now available to provide a reliable assessment of the safety of coastal sand barriers and to reliably predict the flood propagation in coastal areas.

(c) (b)

(a)

11

5 IMPLICATIONS OF COASTAL FLOODS FOR GROUNDWATER

CONTAMINATION: THE CASE OF NEAR BREMERHAVEN

During and after a coastal flood, seawater infiltrates vertically behind overtopped/breached coastal defences, inducing VSWI (Fig 1). In Section 4, XBeach is successfully applied to integrally simulate a barrier breaching/overtopping and the subsequent flood propagation while omitting the infiltration process and the accompanied salinity increase of the originally fresh aquifers. A subsurface model is indeed necessary to simulate coupled groundwater flow as well as advection and dispersion of the seawater in porous media. Therefore, SEAWAT is used to separately simulate the VSWI using the outcomes of XBeach (i.e. water depths and flood extent) as a surface boundary. Hence, the XBeach- SEAWAT system is applied to the case of near Bremerhaven to draw the implications of coastal floods for groundwater contamination.

The case study makes use of available hydro-geophysical data for a 12-km long vertical cross section belongs to the German Bight, which is situated in northern Bremerhaven, northern Germany (Fig 8).

The topography of this profile and the dyke location are shown in Fig (8.e).

Fig. 8: Location and details of the study area near Bremerhaven

(e) (d)

12

The considered storm surge (Fig. 8.d) induces a maximum surge up to 8.5 m.a.s.l., which is about 1.1 m higher than the dyke crest. The surge results in overtopping flow over the dyke crest for 2.8 h. Subsequently, seawater inundates the hinterland behind the dyke and infiltrates vertically into the fresh aquifers. The dyke is considered non-erodible to prevent its overwash by increasing the value of Manning coefficient (𝑛2 = 2.592 𝑚 . 𝑠) in Fig 8.e over the dyke.

Two modelling scenarios for overtopping and propagation are considered (Fig 9): (i) Morpho-off scenario considers that no morphological evolution takes place in XBeach (i.e. sediment transport is omitted), and (ii) Morpho-on scenario permits morphological evolution when the flow velocity exceeds the threshold value for the onset of sediment motion.

(a)

(b)

Fig. 9: Coastal flood propagation: (a) evolution of the bed level (BL) and water levels (WL) and (b) Pre- and post-storm sea, inundation and bed levels at Bremerhaven

-1/3

13

The extremely high Manning value over the dyke zone limits indeed the flow velocity over it to be always (i.e. in both Morpho-on and Morpho-off scenarios) under the threshold value for the onset of sediment motion. As a result, no morphological evolution takes place with both scenarios and hence the inland discharge is identical for both scenarios (= 2,196 m3 with a salt concentration of 2,500 mg/l). As a result, 54.9 tonnes of salt are supplied to the hinterland and then vertically to the aquifers. Fig (9.a) compares the evolution of the bed and water levels for both simulation scenarios, showing that the flood extents are identically increasing with the time marching until water flow is blocked after 10 hours at a cross-shore distance of 6,400 m because of the local increase of the ground elevation. Therefore, the flood extends 5,000 m behind the dyke for both simulation scenarios. Fig (9.b) clearly shows the flood extent through comparing the initial (at t = 0 h) and the final (at t = 10 h) bed and water levels for both simulation scenarios.

The flood extent and water depth from Fig 9 are used as a surface boundary to simulate the VSWI using SEAWAT. Elsayed (2017) reported that the overtopped seawater takes 4 days to infiltrate behind the dyke into the aquifer and to induce a disorder of the salt mass budget in the aquifers due to this coastal flood. The flow directions and salt concentrations in the aquifers are shown in Fig (10) after 1 day, 3 months, 1 year and 20 years, where the salt-freshwater interface is represented by the 50 % iso- concentration contour, and the iso-concentration contour of 500 mg/l (2 %) represents the maximum salt concentration for drinkable water according to the World Health Organization (WHO).

Fig. 10: Salt distribution in Bremerhaven aquifers after 1 day, 3 months, 1 year and 20 years. Arrows represent flow directions.

Before the flood event, only lateral intrusion induced by the hydraulic interconnectivity between sea and groundwater prevails. After flooding, the saltwater infiltrates into the aquifer along the 5-km flood extent and the salt spreads vertically because the infiltrating saltwater is heavier than the prevailing freshwater in the aquifer. Even after 3 months and one year, the salt diffusion is still in the vertical direction. Therefore, saltwater moves vertically beneath the flood extent until it mixes with the fresh water over the entire aquifer depth. Saltwater infiltration deviates seaward under the effect of the regional flow (Fig 1). Such seaward directed flow and infiltrated precipitation generate an effective hydraulic barrier to impede

14

further inland migration of saltwater and provide a downgradient freshwater discharge for saltwater dilution and flushing counteracting the effects of storm surge on the extent of VSWI (as shown by comparing panels of 1 and 20 years in Fig (10)). This dilation process results in a very slow process of natural remediation until the aquifer is remediated totally.

In order to determine the time of full aquifers remediation of a harmless reuse of groundwater as the case before the flood-induced intrusion, Elsayed (2017) computed the three salt budget curves in Fig (11), namely: the accumulative (total) source in mass, the accumulative (total) sink out mass, and the curve of total mass remaining in the aquifer. The latter curve represents the mismatch between the two former curves. The increase of the source in mass during the percolation interval of 4 days (starting from t=1,825 days at the end of the model warming up time to t=1,829 days) is totally stored in the aquifer, as represented by the sudden increase in the curve of the total salt mass remaining in the aquifer. This stored mass sinks out the aquifer gradually until the aquifer is totally remediated after 44.3 years. This indeed highlights how coastal floods might hinder aquifers usability and hence the sustainable development of coastal areas due to long-term recovery. Moreover, crops in hinterlands may suffer stress, thereby not grow properly, or may die due to salt intolerance, thus leading to a decrease in the agricultural yield. Therefore, this study tentatively suggested using an SSDN as a mitigation measure as discussed below.

Fig. 11: Curves of salt budgets in Bremerhaven aquifers. Detail (b) shows the increase of the source in salt mass

owing to flooding.

6 SUBSURFACE DRAINAGE EFFECT ON THE RESILIENCE OF COASTAL

AQUIFERS

Most studies associated with VSWI are limited to the determination of the natural remediation interval; no suitable mitigation measures are proposed to control this type of intrusion and to shorten the commonly long remediation intervals. Fig 12 presents the common strategies for controlling seawater intrusion, which is either in the form of upconing induced by excessive pumping and/or in the form of landward shifting of the salt-freshwater interface due to sea level rise or long-term decline of the GWT.

15

Fig. 12: Common strategies to manage saltwater intrusion

None of these traditional techniques is suitable for managing VSWI. Therefore, Elsayed and Oumeraci (2018) suggested and numerically tested the feasibility of using an SSDN (Fig. 13), especially in flood-prone agriculture areas. The drainage, in general, would absorb the contaminated water before reaching the fresh groundwater. However, surface drainage is inappropriate because it could enlarge the contamination extent and surface drains would act as preferential pathways for landwards movement of seawater.

(a)

(b)

Fig. 13: Illustrations of roles of the subsurface drainage (a) in lowering the GWT and in improving the agricultural yield and (b) in enhancing the resilience of coastal aquifers against coastal floods

Reducing pumping from coastal aquifers

Relocating/shifting extraction wells landward

Directly recharging aquifer (primarily surficial

aquifers), Freshwater recharge into wells paralleling the

coast, creating a hydrodynamic barrier Creating a trough parallel to the coast by

excavating encroaching salt water from wells

Extracting seawater before it reaches wells

Extraction/injection combination

Construction of impermeable subsurface barriers

16

The subsurface drains lower the unwanted increase of the GWT in order to enhance growing of crops (Fig 13.a), thus increasing the agricultural yield. Moreover, the SSDN collects part of the infiltrating saltwater. Nevertheless, the rest is escaping downward among the drains, as schematically shown in Fig (13.b). By comparing both panels of Fig (14) for saltwater distribution in Bremerhaven aquifer for both undrained and drained situations at 1 year after the flood event, it is easy to notice that drainage has confined the high salt concentration near to the ground surface. Elsayed (2017) reported that highly concentrated saltwater is collected from the shallow zones within the three years after flooding and shorter remediation intervals (< 3 years) might be achieved in case of using closer drains. However, the efficient role of the drainage in shortening the remediation time is often at the expense of more lateral intrusion (as shown in Fig (14) by comparing the 50 % iso-concentration contour in panel (a) for undrained and in panel (b) for drained conditions) because of the drainage-induced lowering of the GWT.

Fig. 14: Salt distribution in the aquifer after one year of the flood event (a) with no drainage applied and (b) with subsurface drains

7 SUMMARY, LESSONS LEARNT AND IMPLICATIONS FOR PRACTICAL

APPLICATIONS

A new approach is developed to integrally simulate breaching of CBS, induced inundation and subsequent VSWI. The problem of overestimation of erosion rates and breaching dimensions by XBeach is tentatively solved by introducing two novel physical processes and implementing them in XBeach. Moreover, the scope of XBeach is successively extended to simulate the breaching and the subsequent inundation in a single model. Then the XBeach-SEAWAT approach is applied to the case of near Bremerhaven, showing that a short overtopping event for only 2.8 hours may inundate 5-km and may increase the aquifers salinity for more than four decades. The use of an SSDN as a mitigation measure significantly shortens the natural remediation interval and limits the vertical extent of the contamination. Nevertheless, the latter is often accompanied by an increased lateral intrusion due to the defection in the hydrostatic equilibrium between the mean sea level and the GWT.

Based on these outcomes, the following aspects are recommended for the practical applications:

• The improved XBeach is recommended as a prediction tool to assess vulnerability of coastal defences to surges and to predict inland flow rather than empirical models (e.g. EurOtop).

• Continuous maintenance of the coastal defences is very crucial to avoid flooding and to protect valuable groundwater resources.

• Having highly compacted/consolidated coastal defences that can cope with extreme overtopping without breaching is a crucial issue.

• For the residual inland discharge, it is recommended to install a suitable SSDN.

• Lowering the GWT might be effective in increasing the agricultural yield. However, in coastal areas, it might induce further lateral intrusion.

17

8 ACKNOWLEDGEMENT

I would like to acknowledge the constant guidance and support of my doctoral advisor Professor Hocine Oumeraci. I also thank the NLWKN for providing the GWK-tests. The data of Bremerhaven is provided by Prof Thomas Graf. I also acknowledge the support of Dr Robert McCall (Deltares), who provided data for XBeach validation. Financial support of DAAD for the author's PhD in the frame of the Exceed-Swindon Project is gratefully acknowledged.

9 REFERENCES

Elsayed, S.M. (2017): “Breaching of Coastal Barriers under Extreme Storm Surges and Implications for Groundwater Contamination”, PhD dissertation, Leichtweiß Institute for Hydraulic Engineering and Water Resources, TU Braunschweig, Available at: https://dx.doi.org/10.24355/dbbs.084- 201710161043.

Elsayed, S.M. and Oumeraci, H. (2018): “Modelling and Mitigation of Storm-Induced Saltwater Intrusion: Improvement of the Resilience of Coastal Aquifers Against Marine Floods by Subsurface Drainage”, Environmental Modelling and Software 100, 252-277.

Elsayed, S.M. and Oumeraci, H. (2017): “Effect of Beach Slope and Grain-Stabilization on Coastal Sediment Transport: An Attempt to Overcome the Erosion Overestimation by Xbeach”, Coastal Engineering 121, 179-196.

Elsayed, S.M. and Oumeraci, H. (2016): “Combined Modelling of Coastal Barrier Breaching and Induced Flood Propagation Using Xbeach”, Hydrology 3, 34.

Langevin, C., Jr, D.T., Dausman, A. and Sukop, M. (2008): “SEAWAT Version 4: A Computer Program for Simulation of Multi-Species Solute and Heat Transport”, Virginia.

McCall, R.T., Van Thiel de Vries, J.S.M., Plant, N.G., Van Dongeren, A., Roelvink, J.A., Thompson, D.M. and Reniers, A.J.H.M. (2010): “Two-Dimensional Time Dependent Hurricane Overwash and Erosion Modeling at Santa Rosa Island”, Coastal Engineering 57, 668-683.

Park, W.K., Moon, Y.H., Chang, S.Y., Jeong, W.M., Chae, J.W., Ryu, K.H., Chang, Y.S. and Jin, J.Y. (2018): “Nonlinear Transformation of Storm Waves and Impacts on Nearshore Mound in Haeundae Beach, Korea”, Journal of Coastal Research 85, 1131-1135.

Roelvink, D., Reniers, A., van Dongeren, A., van Thiel de Vries, J., McCall, R. and Lescinski, J. (2009): “Modelling Storm Impacts on Beaches, Dunes and Barrier Islands”, Coastal Engineering 56, 1133-1152.

18

SUMMARY

Europe and many other countries are often surrounded by coastal defences (e.g. protective dunes) in order to protect coastal areas from threats of storm surges and flooding. However, during extreme surges, the higher water levels may temporally threaten these defences. As a result, they may be overtopped/breached, inducing hinterland flooding and subsequent vertical saltwater intrusion (VSWI) behind the breached barriers due to the vertical infiltration of inundating seawater into the fresh groundwater.

In this study1

, a new integral methodology is developed to reliably assess the possible implications of storm surges on the safety of coastal barriers (CBS) and the implications of possible breaching for flood propagation as well as for accompanied contamination of coastal aquifers due to the VSWI. The modelling methodology consists of an improved XBeach code [Roelvink et al., 2009] weakly coupled with the SEAWAT model [Langevin et al., 2008]. XBeach simulates successively breaching of CBS and the subsequent flooding while SEAWAT simulates the VSWI. To achieve reliable modelling of coastal erosion and breaching, some XBeach improvements and extensions are formulated and validated using, among others, unique large-scale dataset for dune erosion. The methodology is then applied to a case study, showing that coastal floods represent a serious threat to coastal aquifers which are extremely important water resources. A flood of a few hours may contaminate coastal aquifers for decades, thus reducing the agricultural yield and hindering the sustainable development in coastal areas. Probably, this is the foremost study that attempts to mitigate storm-induced VSWI through the use and modelling of a subsurface drainage network (SSDN). Besides improving the agricultural yield, the use of an SSDN significantly shortened the natural remediation interval required for aquifers recovery. The multiple flow domains make this study quite relevant for the coastal engineering community, for flood risk managers, for groundwater suppliers as well as for sustainable development planners.

1 The presented study was conducted within the author's doctoral studies at the Leichtweiß-Institute from June 2014 to August 2017.

This article summarises the main findings of the author’s doctoral dissertation [Elsayed, 2017], which is freely available under

https://doi.org/10.24355/dbbs.084-201710161043. The main outcomes of this dissertation are published in three journal papers

[Elsayed and Oumeraci, 2018, 2017, 2016]. Moreover, they are discussed in 4 international conferences (the ICCE 2018, the USA;

the 5th IAHR Europe congress, Italy; The XBeach X conference, the Netherlands; the INECEP summer school, Mexico).

Furthermore, details of this dissertation are available in five technical reports which are also freely available on the author’s page

on ResearchGate (https://www.researchgate.net/profile/Saber_Elsayed2)

RESUME L'Europe et de nombreux autres pays sont souvent entourés de défenses côtières (par exemple, des dunes de protection) afin de protéger les zones côtières contre les menaces de marées de tempête et d'inondations. Toutefois, lors de marées extrêmes, les niveaux d'eau plus élevés peuvent temporairement menacer ces défenses. En conséquence, elles peuvent être débordées ou percées, provoquant des inondations dans l'arrière-pays et l'intrusion verticale d'eau salée (VSWI) derrière les barrières percées en raison de l'infiltration verticale de l'eau de mer inondante dans la nappe phréatique douce.

Dans cette étude1, une nouvelle méthodologie intégrale est développée pour évaluer de manière fiable les implications possibles des ondes de tempête sur la sécurité des barrières côtières (CBS) et les implications d'une brèche éventuelle pour la propagation des inondations ainsi que pour la contamination accompagnée des aquifères côtiers due à la VSWI. La méthodologie de modélisation consiste en un code XBeach amélioré [Roelvink et al., 2009] faiblement couplé au modèle SEAWAT [Langevin et al., 2008]. XBeach simule successivement la rupture du CBS et l'inondation qui s'ensuit, tandis que SEAWAT simule la VSWI. Pour obtenir une modélisation fiable de l'érosion côtière et des brèches, certaines améliorations et extensions de XBeach sont formulées et validées en utilisant, entre autres, un ensemble de données unique à grande échelle pour l'érosion des dunes. La méthodologie est ensuite appliquée à une étude de cas, montrant que les inondations côtières représentent une menace sérieuse pour les aquifères côtiers qui sont des ressources en eau extrêmement importantes.

19

Une inondation de quelques heures peut contaminer les aquifères côtiers pendant des décennies, réduisant ainsi le rendement agricole et entravant le développement durable dans les zones côtières. Il s'agit probablement de la principale étude qui tente d'atténuer les inondations induites par les tempêtes grâce à l'utilisation et à la modélisation d'un réseau de drainage souterrain (SSDN). Outre l'amélioration du rendement agricole, l'utilisation d'un réseau de drainage souterrain a permis de réduire considérablement l'intervalle d'assainissement naturel nécessaire à la récupération des aquifères. Les multiples domaines d'écoulement rendent cette étude tout à fait pertinente pour la communauté du génie côtier, pour les gestionnaires des risques d'inondation, pour les fournisseurs d'eaux souterraines ainsi que pour les planificateurs du développement durable.

1 L'étude présentée a été menée dans le cadre des études doctorales de l'auteur à l'Institut Leichtweiß de juin 2014 à août 2017. Cet article résume les principales conclusions de la thèse de doctorat de l'auteur [Elsayed, 2017], qui est disponible gratuitement sur https://doi.org/10.24355/dbbs.084-201710161043. Les principaux résultats de cette thèse sont publiés dans trois articles de revue [Elsayed et Oumeraci, 2018, 2017, 2016]. De plus, ils sont discutés dans 4 conférences internationales (le CIEC 2018, Etats-Unis ; le 5ème congrès européen de l'AIHR, Italie ; la conférence XBeach X, Pays-Bas ; l'université d'été de l'INECEP, Mexique). De plus, les détails de cette thèse sont disponibles dans cinq rapports techniques qui sont également disponibles gratuitement sur la page de l'auteur sur ResearchGate (https://www.researchgate.net/profile/Saber_Elsayed2)

ZUSAMMENFASSUNG

Die europäischen und viele andere Länder sind oft von Küstenschutzbauwerken umgeben (z. B. schützende Dünen), um Küstenregionen vor der Bedrohungen durch Sturmfluten und Überflutungen zu schützen. Während extremer Fluten können höhere Wasserstände diese Schutzvorrichtungen jedoch zeitweise gefährden. Infolge dessen können sie überflutet oder beschädigt werden, was eine Flutung des Hinterlandes verursacht und eine vertikale Salzwasserintrusion (VSWI) hinter den gebrochenen Absperrungen durch die vertikale Infiltration von überflutendem Salzwasser in das Grundwasser zu Folge haben kann.

In dieser Studie1 wird eine neue integrale Methode entwickelt, um die möglichen Auswirkungen von Sturmfluten auf die Sicherheit von Küstenschutzbauwerke (coastal barriers, CBS), die Auswirkungen der Flutausbreitung bei deren möglichen Beschädigung und die damit verbundene Kontaminierung des Aquifers bedingt durch VSWI verlässlich einschätzen zu können. Die Modellierungsmethode besteht aus einem verbesserten XBeach Code [Roelvink et al., 2009] der mit dem SEAWAT Modell [Langevin et al., 2008] schwach gekoppelt wurde. XBeach simuliert sukzessive die Beschädigung von CBS und die folgende Überflutung, während SEAWAT die VSWI simuliert. Um eine verlässliche Modellierung der Küstenerosion und der Beschädigung zu erzielen, werden einige Verbesserungen und Erweiterungen von XBeach formuliert und validiert, indem, unter anderem, einmalige großmaßstäbliche Datensätze für Dünenerosion verwendet werden. Die Methode wird dann in einer Fallstudie angewendet, die zeigt, dass Überflutungen von Küstengebieten eine ernsthafte Bedrohung für Küsten-Aquifere und damit für eine extrem wichtige Wasserressource darstellen. Eine nur wenige Stunden andauernde Überflutung kann Küsten-Aquifere über Jahrzehnte kontaminieren und somit den landwirtschaftlichen Ertrag verringern und eine nachhaltige Entwicklung der Küstenregionen behindern. Wahrscheinlich ist dies die wichtigste Studie, die versucht, sturminduzierte VSWI durch die Anwendung und Modellierung eines unterirdischen Entwässerungsnetzes (subsurface drainage network, SSDN) abzumildern. Neben der Verbesserung des landwirtschaftlichen Ertrags verkürzte die Anwendung eines SSDN deutlich den natürlichen Wiederherstellungszeitraum der für die Erholung des Aquifers erforderlich ist. Die verschiedenen Abflussbereiche machen diese Studie sehr relevant für das Küsteningenieurwesen, für Manager von Hochwasserrisiken, für Grundwasserversorger sowie für Personen in der nachhaltigen Planung.

1 Die hier präsentierte Studie wurde im Rahmen der Dissertation des Autors am Leichtweiß-Institut von Juni 2014 bis August 2017 durchgeführt. Dieser Artikel fasst die wesentlichen Ergebnisse der Dissertation des Autors [Elsayed, 2017] zusammen, welche unter https://doi.org/10.24355/dbbs.084-201710161043 frei verfügbar ist. Die wichtigsten Ergebnisse dieser Dissertation wurden in drei Zeitschriften-Artikeln veröffentlicht [Elsayed and Oumeraci, 2018, 2017, 2016]. Darüber hinaus wurden sie auf vier internationalen Konferenzen diskutiert (ICCE 2018, USA; 5. IAHR Europa-Kongress, Italien; XBeach X Konferenz, Niederlande; INECEP Summer School, Mexiko). Außerdem stehen die Einzelheiten dieser Dissertation in fünf technischen Berichten zur Verfügung, die über die Homepage des Autors bei ResearchGate, ebenfalls kostenlos, zur Verfügung stehen (https://www.researchgate.net/profile/Saber_Elsayed2).

20

RESUMEN

Determinados países en Europa y en otras regiones disponen en muchas ocasiones de defensas costeras (por ejemplo, sistemas dunares) a modo de protección frente a las amenazas derivadas de la acción de los temporales y frente a posibles inundaciones. Sin embargo, durante sucesos extremos, el nivel de agua puede llegar a superar estas defensas. Como resultado, se pueden producir situaciones de rotura/rebase que generan inundaciones en su zona de trasdós y una consiguiente intrusión de agua salina que puede contaminar los acuíferos de agua dulce por efecto de la infiltración en el terreno.

En este estudio se desarrolla una nueva metodología integral para una gestión confiable de estos activos, particularmente en lo que se refiere a las posibles situaciones derivadas por fenómenos de rebase sobre la seguridad de las estructuras de protección, así como las implicaciones de situaciones de inundación y contaminación de acuíferos de agua dulce. La metodología de modelado se basa en un código “XBeach” (Roelvink et al., 2009) mejorado, acoplado al modelo “SEAWAT” (Langevin et al., 2008). El modelo “XBeach” simula sucesivas roturas de las defensas costeras con los consiguientes fenómenos de inundación asociados, mientras que el modelo “SEAWAT” simula las condiciones de intrusión salina. Para alcanzar resultados fiables, algunas de las mejoras y ampliaciones del modelo “XBeach” se han formulado y validado usando, entre otros, datos derivados de modelos a gran escala del comportamiento erosivo de los sistemas dunares. La metodología ha sido aplicada a un caso práctico, mostrando que las inundaciones costeras representan una seria amenaza para los acuíferos ubicados en zonas costeras, que pueden ser un recurso importante para el suministro de agua. Una inundación de varias horas de duración puede llegar a contaminar acuíferos durante décadas, con la consiguiente reducción de la producción agrícola de la zona y la puesta en riesgo del desarrollo sostenible de las áreas costeras afectadas. Posiblemente, estamos ante el principal estudio que tiene como objetivo la reducción de las condiciones de intrusión salina, utilizando para ello técnicas de modelado para poder definir sistemas de drenaje profundo. Además de la mejora que puede suponer para la producción agrícola, el uso de sistemas de drenaje permitiría reducir significativamente los periodos necesarios para la recuperación de las condiciones naturales de los acuíferos. Todo lo anterior hace de este estudio un elemento relevante desde diversos puntos de vista; la ingeniería de costas, la gestión de inundaciones, el suministro de agua, así como para una planificación sostenible del territorio.

1 El estudio presentado se realizó dentro de los estudios de doctorado del autor en el Instituto Leichtweiß de junio de 2014 a agosto de 2017. En este artículo se resumen las principales conclusiones de la tesis doctoral del autor [Elsayed, 2017], que se puede consultar gratuitamente en https://doi.org/10.24355/dbbs.084-201710161043. Los principales resultados de esta tesis se publican en tres artículos de revista [Elsayed y Oumeraci, 2018, 2017, 2016$. Además, se examinan en cuatro conferencias internacionales (el ICCE 2018, EE.UU.; el 5º congreso europeo de la AIDH, Italia; la conferencia XBeach X, Países Bajos; la escuela de verano del INECEP, México). Además, los detalles de esta disertación están disponibles en cinco informes técnicos que también están disponibles gratuitamente en la página del autor en ResearchGate (https://www.researchgate.net/profile/Saber_Elsayed2).

21

Crane barge

Hook

S

PIANC De Paepe-Willems Award 2019 – Shared Second Place

PREDICTION OF SHACKLE MOTION HANGED FROM A JIB TOP OF CRANE BARGE BY A COUPLING NUMERICAL MODEL OF THREE MOTIONS

by

YOSHINOSUKE KURAHARA, M. Eng.

Researcher, Coastal and Ocean Engineering Group, Research and Development Center, TOA CORPORATION

1-3 Anzen-cho, Tsurumi-ku, Yokohama City, Kanagawa, 230-0035, Japan.

E-mail: y_kurahara@toa-const.co.jp

MASAHIDE TAKEDA, P.E.Jp (Civil Eng.), Dr. Eng.

Group Manager, Coastal and Ocean Engineering Group, Research and Development Center, TOA CORPORATION

HAJIME MASE, Dr. Eng.

Technical Adviser, TOA CORPORATION Research Professor, Disaster Prevention Research Institute, Kyoto University

Keywords: hook, shackle, crane barge, oscillation prediction, oscillation control, numerical model

Mots clés : crochet, manille, barge-grue, prédiction des oscillations, contrôle des oscillations, modèle numérique

1 INTRODUCTION

Crane barge, shown in Figure 1, is an indispensable vessel for marine construction works and various cargo handling operations at sea. When handle cargoes using a crane barge, the hull and suspended loads are usually shaken due to waves. The prediction and control of the shaking or oscillation are extremely important from the viewpoint of safety operation, increase the effective working days, accurate construction work and so on. To predict the oscillation, Nojiri and Mita (1980) have developed a computation method for coupled motions of crane barge and suspended load based on a linear theory. They have found that the developed method is able to explain the characteristics of coupled motions by comparing the predictions with the experimental results using a 1/50 scale model of 2,500 tonnes crane barge.

Jib top

hackle

Figure 1: Sketch of crane barge

22

Before lifting the suspended loads, slinging work has to be done; that is hanging the load on hook using a hooking tool such as wire ropes. In the case of crane barge, a U-shaped shackle is attached to the tip of a wire hanging from the hook in advance. Then the shackle is connected to the lifting lug welded onto the load for slinging work. The shackle weight for a load of 100 tonnes is from 150 to 250 kg for the crane work when using a large crane barge. In the open ocean, these large shackles swing with large amplitudes as shown in Figure 2. Even in calm sea conditions where the significant wave height is 0.5 m or less, the workers cannot catch the shackles and cannot continue the crane work. In this way, the operation rate of construction work is greatly affected depending on whether the slinging can be done or not. Therefore, since the shaking prediction of shackle is important, we developed and proposed a numerical model for coupling of double pendulum motions of hook, shackle and ship motions based on a linear theory, especially to predict the motion of shackle on a crane barge.

Figure 2: Swinging shackles during slinging work

2 NUMERICAL MODEL

2.1 Coupling Model of Motions and Coordinate System

The numerical model proposed here to predict the motion of crane barge is a radiation/diffraction panel model based on the linear theory. The model considers the interaction between surface waves and crane barge is based on a three-dimensional panel method.

It is assumed in this model that waves and hull motions are small, and the infinite domain is analysed by the linear theory. The fluid is assumed to be non-viscous, incompressible and irrotational motion being described by a velocity potential. A crane barge is modeled as a box-shaped three-dimensional rigid body. In the Cartesian coordinate system (X, Y, Z), as shown in Fig. 3, the X and Y axes are set on the still water surface and the Z axis is in the vertical downward direction, and the center of the hull is set as the origin O. The translational motion in each axis direction is denoted as X1 for surge, X2 for sway, X3 for heave, and the rotational motion around each axis is denoted as X4 for roll, X5 for pitch and X6 for yaw. The hook and the shackle are suspended from the jib top located at (lx, ly, lz). A hook's x direction motion in the XZ

plane is denoted as X7, shackle’s x direction motion as X8, in the YZ plane X9 is for hook’s y direction motion and X10 for shackle’s y direction motion. Thus, the motion modes of crane barge, hook and shackle are denoted as Xj (j = 1 to 10). The β is the angle between the incident direction wave and the positive X axis as defined in Fig. 3.

23

X

Figure 3: Coordinate system

2.2 Governing Equation

2.2.1 Double Pendulum Motions of Hook and Shackle

The hook and shackle of a crane barge show double pendulum motions fixed at a jib top. The motions of the hook and shackle are caused by motion of the jib top. The motions of hook and shackle in the XZ

plane are shown in Figure 4. The length of the hanging from the jib top to hook is l7, the length of the hanging from hook to shackle l8, and the length of hanging from the jib top to shackle l (=l7+l8). The hook weight is m1, shackle weight m2, and the total hanging weight m (=m7+m8). In the XZ plane, the horizontal displacement and swing angle of hook are X7 and φ7, those of shackle X8 and φ8, and the horizontal displacement of the jib top Xj-xz. Relationship between the horizontal movement distance of the hook and the shackle and the swing angle is expressed by the formula of Eq. (1).

xj-xz

JIB TOP

HOOK

m 8

SHACKLE

Figure 4: Double pendulum model consist of hook and shackle

(1)

Crane jib top

( lx, ly, lz)

lz

lx

Hook

ly

Shackle

X1

X4

X2 X3

X5 X6

Incident wave

φ7

X7 m 7

X φ7

l7

l8

24

The motions of the hook and shackle in the XZ plane can be expressed by the Lagrange equation of Eq. (2). The determinants A, B, C and D are given by Eq. (3).

where c7 and c8 are the damping coefficients related to the damping force term proportional to the horizontal displacement speed. The displacement of the jib top can be expressed by Eq. (4) derived from the movement of crane barge.

Similarly, the motion of hook and shackle in the YZ plane can be calculated by the simultaneous differential equations of hook X9 and shackle X10. In the YZ plane, it is sufficient to give the displacement Xj-yz of the jib top as Eq. (5) from the movement of crane barge.

2.2.2 Analysis Method of Hull Motion

Generally, wave-exciting force (Froude-Krylova force + diffraction force), radiation force, and static restoring force act on a floating body moving in waves. In addition to these forces, a coupled force acts on the jib top of crane barge as a dynamic reaction force due to the double pendulum motions of hook and shackle. Therefore, the equation of motions of crane barge is expressed as Eq. (6).

(6)

where FDi(t) : the wave-exciting force, FRi(t) : the radiation force, FSi(t): the static restoring force, FCi(t) : the coupled force by hook and shackle double pendulum motions. In this model, wind drag force, flow drag force, mooring force and other environmental external forces are not considered. Wave-exciting force and radiation force are calculated from three-dimensional velocity potential using the three-dimensional singularity distribution method [Tsutsumi et al., 1974 ; Inglis and Price, 1980]. The static restoring force is calculated from the balance between the centre of hull’s gravity and buoyancy forces. 1) : Wave-exciting force

As shown in Figure 5, the force (wave-exciting force and radiation force) that the hull of crane barge receives from the water surface is calculated by adding the fluid fluctuating pressure p(P, t) acting on the point P(xp, yp, zp) in the total surface area SH. As shown in Eq. (7), the fluid fluctuation pressure of waves is calculated by the speed potential with a normal vector ni.

25

( ) = − ( , ) ( ) = ρ Φ ( , ) ( ) (7)

Figure 5: Discretisation of hull under the water surface to a microscopic plane element

The velocity potential ΦD (P,t) related to wave-exciting force is given by Eq. (8); ΦD (P,t) is the sum of the incident wave velocity potential Φ0 (P,t) and the velocity potential Φ7 (P,t) of the fluid dispersion resulting from the hull reflection disturbance.

2) : Radiation force Similarly, the radiation force is expressed as Eq. (9) using the velocity potential.

The velocity potential ΦR(P,t) of the radiation force is the sum of velocity potential Φj(P,t) caused by waves consisting of motion mode (j = 1 to 6) of the hull of crane barge, described as Eq. (10).

The velocity potential Φj (P,t) fluctuates with time. Specifically, it is calculated from Eq. (11) by the convolution operation of the impulse velocity potential ΔΦj (P,t) by an unit velocity and the motion speed

Xj(t) from the past to present time t as follows:

Φ ( , ) = ∆Φ ( , − ) ̇ ( ) (11)

Impulse velocity potential ΔΦj(P,t) by an unit velocity is calculated by Eq. (12) [Takagi and Arai, 1996] bellow.

where, δ(t) : the delta function (δ(t)=0 for t≠0, δ(t)=∞ for t=0), H(t) : Heaviside function (H(t)=1 for t≥0, H(t)=0

for t<0), Ωj(P) : the velocity potential of turbulent wave in the vicinity of the hull, Γj(P,t) : the velocity potential of divergent wave from the hull in far of the hull.

.

26

Therefore, the radiation force is given by Eqs. (13) and (14).

where, mij(∞) : the additional mass coefficient at ω=∞ (generally not zero), Lij(t) : the memory influence function of fluid force.

3) : static restoring force

As shown in Eqs. (15) and (16), the static restoring force is expressed using the modes ij = 33, 35, 53, 44, and 55. The other modes of Cij are zero.

where, AW is the area of waterline surface, GMB the horizontal meta center height, GML the vertical meta

center height, ∇ the displacement, as shown in Fig. 6.

Figure 6: Meta center height appeared in Eq. (6)

27

4) : Coupled force by hook and shackle motions

The coupled force due to the double pendulum motions of hook and shackle act on the jib top of crane barge. The coupled force is expressed by Eq. (17).

2.2.3 Cross-Coupling Force

The equation of motions of crane barge, hook, and shackle is summarised in a form of the second-order linear differential equations as Eq. (19).

where, mij is the generalised mass, mij(∞) the additional mass when frequency is infinite, Lij(t) the memory influence function, Cij the restitution coefficient, Ei(t) the wave-exciting force. Since the hook and shackle constitute a double pendulum, the hook and shackle influence each other and induce irregular response even if regular external forces are acted. Therefore, the coupling equation as a function of time must be solved as an initial value problem.

First, Lij(t) on the left side of Eq. (19) can be obtained from Fourier transformation of the wave damping coefficient Bij(t).

Next, Ei(t) is obtained by solving the integral equation based on the Green function in the time domain for the velocity potential under the boundary condition. The calculation of the Green function in the time domain requires a large amount of storage capacity and computation time; it is not suitable for analysing the influence of the shape parameter of crane barge by changing its value. Therefore, in this study, by using the radiation velocity potential obtained in the process of calculating the memory influence function, the wave-exciting force is calculated by the Haskind relation [Takagi and Arai, 1996] from the radiation velocity potential.

Assuming that the impulse response function is ei(t) and the time series of the wave height is h(t), the wave-exciting force Ei(t) becomes Eq. (21).

Here, h(t) is the water surface variation of irregular wave estimated from the wave spectrum. Eq. (19) describes the equation of motions of crane barge at the origin O shown in Fig. 3; that is, h(t) is the water surface variation at the origin. In this analysis, the irregular wave spectrum is given as Bretschneider spectrum.

28

The impulse response function, ei(t), is obtained from the Fourier transform of the response function Hi(t)

of the wave-exciting force as follows.

3 HYDRAULIC EXPERIMENTS

3.1 Hydraulic Experiments of Crane Barge Motion [Nojiri and Mita, 1980]

First, the verification of the present coupling motion model of the crane barge, hook and shackle was carried out using the results obtained by Nojiri and Mita (1980). They measured the frequency response of a crane barge motion in a two-dimensional wave flume. Table 1 shows the main specifications of the experiments.

The suspended load in their experiment is treated as a hook weight. Figures 7 and 8 show the comparison of the roll of crane barge and the hook swing angle respectively. The horizontal axis is the dimensionless frequency of the incident wave using the ship width B and gravity g, the vertical axis is the dimensionless amplitude of the hook swing angle divided by the wave slope (ka: k is the wave number, a is the wave amplitude). It is seen from the figures that the present numerical model gives good predictions of coupled motions of the crane barge and hook for roll and swing against all range of frequency with slight difference at the peak values.

Item Symbol BARGE A BARGE B

hull length L (m) 2.470 2.470

hull width B (m) 0.800 0.500

draft d (m) 0.100 0.200

jib top coordinates

lx (m) 0 0

ly (m) 0.425 0.425

lz (m) 0.750 0.650

hook weight m (kg) 6.000 6.000

length of hanging l (m) 0.408 0.300

Table 1: Main specifications of experiments

Figure 7: Comparison of the roll of crane barge

29

Figure 8: Comparison of the hook swing angle

3.2 Experiments of Double Pendulum Motions

Next, to verify the present numerical model of the double pendulum motions of hook and shackle, laboratory experiments using an exciter was performed. A conceptual diagram and phots of the laboratory experiment are shown in Figure 9.

In these experiments, the horizontal displacement of the jip, which is a fulcrum of a pendulum, was provided by an exciter (SSV-125, SANESU). The swing response of hook and shackle was measured. The motions of hook and shackle were measured with two video cameras (SSC-DC690, SONY) where the motions of light reflective markers attached to hook and shackle were taken and converted into displacement by image analysis.

Figure 9: Conceptual diagram and phots of the laboratory motion experiments

30

The scale of experiments was 1/20, and the scaling was done according to Froude law. The exciter motion of the jib top was changed as two kinds of waveforms: a sinusoidal waveform and an irregular waveform exerted 50 waves or more. There were two experimental cases with different hanging lengths, as shown in Table 2. CASE-A is a reproduction of the construction work condition using a steel pipe pile by crane barge. Slingers connecting hook and shackle from jib top used in this experiment were synthetic polyvinylidene fluoride with low elongation at break.

The time series of motions of hook and shackle are shown in Figures 10 and 11, respectively, when the jib top is subjected to sinusoidal excitations of 8 s period in the prototype scale. The vertical axis is the dimensionless swing amplitude normalised by the amplitude of the jib top motion. The predicted results agree well with the experimental data. The swing of hook has two peak periods of 8 s and 16 s in Figures 10 and 11. Experimental motions seem to have a very long period trend compared to the calculated ones; this phenomenon is considered to be due to the influence of the longitudinal sling's vibration in the longitudinal direction, although slingers with low elongation at break were used. The calculated value of shackle has the same swing amplitude and period trend as the experimental value; however, the amplitude is slightly smaller than the experimental value. In the calculation, the damping coefficient c8 = 0.05 of the sling, determined from the measured swing amplitude of the shackle of the actual machine in sea waves, is considered. However, in the experiment, the shackle weight is m8

= 0.003 kg and the estimated value of the damping coefficient c8 is somewhat larger.

CASE-A CASE-B

model actual model actual

hook length of hanging l7 (m) 3.25 65.0 4.00 80.0

weight m7 (kg) 1.3 10,000 1.3 10,000

shackle length of hanging l8 (m) 1.25 25.0 0.50 10.0

weight m8 (kg) 0.003 20.0 0.003 20.0

Table 2: Experimental cases of hanging length in protptype scale

Figure 10: Time series of hook motion in regular excitation (CASE-A)

Figure 11: Time series of shackle motion in regular excitation (CASE-A)

31

Figure 12 shows a comparison between experimental and calculated significant swing amplitudes of hook and shackle when irregular excitation is applied to the jib top. The excitation direction of the jib top, simulating waves, is 45° with respect to the ship axis. The horizontal excitation displacement is given as 1 m corresponding to the incident significant wave height. The horizontal axis of Fig. 12 is the incident significant wave period, and the vertical axis the significant swing amplitude Xs of the hook and shackle normalised by the significant excitation amplitude Xj of the jib top.

It is seen that in CASE-A the shackle oscillates around the hook with large amplitude of which ratio is from 8 to 10 at any periods. Those phenomena are well predicted by the present numerical model. It is also seen that the dimensionless hook and shackle motions are small in the range from 1.0 to 2.0, and predictions agree well with the experimental values. By checking all experimental and calculated results, the validity of the present numerical model was ascertained for the crane barge motion and double pendulum motions of hook and shackle.

Figure 12: Dimensionless significant oscillation amplitude of hook and shackle in irregular excitation

4 DISCUSSION

The proposed and validated numerical model was employed to investigate the oscillation characteristics of a crane barge, hook and shackle in waves. The analysis was conducted on a 1,600 tonnes crane barge (prototype size: L = 106 m, B = 43 m, d = 4.35 m). The hook and shackle were assumed to be suspended from the jib top of which position (lx, ly, lz) = (50 m, 30 m, -95 m), and their weights were set to m7 = 10 tonnes and m8 = 0.02 tonnes. Firstly, the frequency characteristics of jib top, hook and shackle motions in waves were investigated by spectral analysis. Figure 13 shows the frequency spectra of jib top, hook and shackle motions in CASE-1 where the hook is located 25 m above the shackle and waves attack from 45° angle to the ship’s X axis. The significant wave height and period are 1 m and 10 s, respectively. The natural oscillation periods of hook (T7) and shackle (T8) are summarized in Table 3, assuming that the righthand side of Eq. (2) is zero. As the position of the hook becomes lower, T7 becomes short, and conversely, T8

becomes long. The jib-top motion spectrum has a peak around 0.1 Hz corresponding to the incident wave spectrum. On the other hand, the hook motion spectrum has a peak around 0.06 Hz on the low frequency side. The shackle motion spectrum has a sharp peak at 0.1 Hz and another peak around 0.06 Hz. Thus, it can be seen that each motion has different frequency characteristics.

32

Figure 13: Frequency spectrum of jib top, hook, shackle motion

Subsequently, we examined the characteristics of shackle motion under the conditions that the position of shackle was the same (constant) and the hook was changed its position from 25 m to 5 m above the shackle at an interval of 5 m (denoted as CASE-1 to CASE-5). Figure 14 shows the dimensionless significant amplitude of shackle motion against the wave period when irregular wave incidence is 45° from the ship axis. In the case that the hook is as high as 25 m above the shackle (CASE-1), the shackle is severely shaken by waves with 10 s period. The significant amplitude of this motion is 17 m. On the other hand, in the case of hysteresis of hook as low as 5 m above shackle (CASE-5), the oscillation is small, and the significant amplitude is 1 m. In CASE-1 to CASE-3, the peak of the shackle's significant amplitude largely appears to be around T7. On the other hand, in CASE-4 to CASE-5, the peak of the shackle's significant amplitude appears to be on the long period around T8. For waves with a wide range of periods seen in field ocean, if the natural periods T7 and T8 are close to each other, there is a high possibility that the oscillation of shackle will increase synchronously with the wave period. Therefore, if the distance between the hook and the shackle can be set so that the interval between T1 and T2 is large, the oscillation of the shackle can be reduced.

CASE-1 CASE-2 CASE-3 CASE-4 CASE-5

T7 (s) 10.02 8.96 7.76 6.34 4.48

T8 (s) 16.19 16.80 17.39 17.95 18.51

Table 3: Natural period of hook and shackle

Figure 14: Dimensionless significant oscillation amplitude of shackle by changing hook position

33

Based on this finding, we actually measured the shaking motion by changing the distance between the

hook and shackle from 6 m to 2 m, using a 200-tonne crane barge as shown in Figure 15. It is found that

the significant amplitude of shackle oscillation could be reduced from 4.2 m to 1.9 m, which turns to be

a 55-% reduction.

Figure 15: Field observation of hook and shackle motion hanged from a jib top of crane barge

34

5 CONCLUSIONS

We have developed a numerical model to analyse the cross-coupled motion between a crane barge in waves and a double pendulum consisted of hook and shackle. A set of experimental data was used to validate the model. The validated model is able to reproduce the swing motion of hook on a crane-barge. One of the objectives of the present study is how to reduce the swing motion of shackles using the numerical simulation model. The important conclusions of this study are summarised as follows:

1) The motion of shackle on a crane barge can be well predicted using the coupling model of motions proposed in this study.

2) The motion of shackle on a crane barge is larger than that of hook, and the motion of shackle is important in slinging work. The success or failure of slinging work significantly influences the rate of effective working days of maritime construction work.

3) When performing the slinging work, the swinging amplitude of shackle can be reduced by shortening the distance between the hook and shackle. This finding is especially important for the safety of marine construction work.

We believe that application of this research makes the development of accuracy and safety technologies in actual marine crane work.

6 ACKNOWLEDGMENTS

The authors would like to thank to Dr. Hiroshi Kawabe belonging to Ship and Ocean Engineering Consultant Limited Liability Company for assistance with the numerical simulations and laboratory experiments.

7 REFERENCES

Nojiri, N. and Mita, S. (1980): “On the Coupled Motions between a Crane Barge and Hook Load in

Wave”, Transactions of the West-Japan Society of Naval Architects, Vol. 59, pp. 43-55.

Tsutsumi, T., Ogiwara, S. and Jinnaka, T. (1974): “On the Principal Particulars of Ship Hull Form and

Wave Pattern Resistance (I)”, Journal of the Society of Naval Architects of Japan, Vol. 136, pp. 17-27.

Inglis, R. B. and Price, W. G. (1980): “Comparison of Calculated Responses for Arbitrary Shaped Bodies using Two and Three-Dimensional Theories”, International Shipbuilding Progress, Vol.27, No. 308, pp. 86-95. Takagi, M. and Arai, S. (1996): “Wave Resistance Theory of Ships and Marine Structures”, Seizando-Shoten publishing co.ltd, pp. 590-593.

35

Crane barge

Hook

S

SUMMARY

Crane barge, shown in Fig. 1, is an indispensable vessel for marine construction works and various cargo handling operations at sea. When handle cargoes using a crane barge, the hull and suspended loads are usually shaken due to waves. The prediction and control of the shaking or oscillation are extremely important from the viewpoint of safety operation, increase the effective working days, accurate construction work and so on. To predict the oscillation, Nojiri and Mita (1980) have developed a computation method for coupled motions of crane barge and suspended load based on a linear theory. They have found that the developed method is able to explain the characteristics of coupled motions by comparing the predictions with the experimental results using a 1/50 scale model of 2,500 tonnes crane barge.

Jib top

hackle

Figure 1: Sketch of crane barge

RESUME

La barge-grue, illustrée dans la Figure 1, est un navire indispensable pour les travaux de construction maritime et les diverses opérations de manutention de cargaisons en mer. Lorsqu'on manipule des cargaisons à l'aide d'une barge-grue, la coque et les charges suspendues sont généralement secouées par les vagues. La prévision et le contrôle des secousses ou des oscillations sont extrêmement importants du point de vue de la sécurité des opérations, de l'augmentation du nombre de jours de travail effectif, de la précision des travaux de construction, etc. Pour prédire l'oscillation, Nojiri et Mita (1980) ont mis au point une méthode de calcul des mouvements couplés de la barge de la grue et de la charge suspendue basée sur une théorie linéaire. Ils ont découvert que la méthode développée est capable d'expliquer les caractéristiques des mouvements couplés en comparant les prévisions avec les résultats expérimentaux en utilisant un modèle à l'échelle 1/50 de 2.500 tonnes de barge-grue.

ZUSAMMENFASSUNG

Ein Kranschiff (Bild 1) ist ein unverzichtbares Schiff für Bauarbeiten im marinen Bereich sowie für den Ladungsumschlag auf See. Wenn beim Umschlag von Ladungen Kranschiffe verwendet werden, werden der Schiffskörper und die schwebenden Lasten gewöhnlich bedingt durch Wellen erschüttert. Die Vorhersage und Kontrolle der Erschütterung oder Schwingung ist extrem wichtig für einen sicheren Betrieb, zur Erhöhung der Effizienz der Arbeitstage, für ein exaktes Bauen, usw. Um die Schwingung vorherzusagen, haben Nojiri and Mita (1980) eine Berechnungsmethode für die gekoppelten Bewegungen des Kranschiffs und der schwebenden Last entwickelt, die auf einer linearen Theorie

36

basiert. Sie haben herausgefunden, dass die entwickelte Methode in der Lage ist, die Charakteristiken der gekoppelten Bewegungen zu beschreiben, indem sie die Vorhersagen mit den Versuchs-ergebnissen unter Verwendung eines Modells eines 2.500 Tonnen schweren Kranschiffs im Maßstab 1 : 50 vergleichen.

RESUMEN

Las barcazas dotadas de grúa (como la que se muestra en la Figura 1) son elementos indispensables para el desarrollo de las obras marítimas y para efectuar operaciones de manipulación de cargas a flote. Cuando se producen operaciones con barcaza, tanto el casco de la misma como la propia carga suspendida sufren movimientos debido a la acción del oleaje. La predicción y control de este tipo de situaciones son extraordinariamente importantes para garantizar unas adecuadas condiciones de seguridad, permitir una mejora de la productividad de las operaciones, así como garantizar la correcta precisión en el desarrollo de las mismas. Para predecir estas oscilaciones, Nojiri and Mita (1980) han desarrollado un método computacional acoplando los movimientos de la barcaza y de la carga suspendida, utilizando para ello una teoría lineal. Se ha concluido que el método desarrollado es capaz de explicar los movimientos acoplados si se compara con los resultados experimentales obtenidos a través de modelos a escala 1/50 de una barcaza de 2.500 t.

37

PIANC De Paepe-Willems Award 2019 – Shared Second Place

ASSESSMENT OF OVERTOPPING VERTICAL RIVER WALLS

DUE TO VESSEL-GENERATED WAVES (VESSEL WASH)

by

Javier Murgoitio Esandi

Ports & Marine Team, AECOM, 2 Leman St, London LE1 8FA; email:

Javier.MurgoitioEsandi@aecom.com

Keywords: vessel wash, overtopping, vertical wall

Mots clés : lavage de la cuve, débordement, paroi verticale

1 INTRODUCTION

Riverside walkways or promenades are very common in towns and cities where the waterway is used by ships carrying freight or passengers and pleasure boats for recreation. Although they may not be regularly flooded due to raised water levels, members of the public standing, walking or cycling in such areas may still be exposed to the hazard of overtopping due to waves from vessel wash. This is a consideration not just for new waterside development, but also existing structures on the banks of rivers that may be more exposed over time with sea level rise.

The widely used guide to assess overtopping is EurOtop (2007). It is based on considerable data gathered by the European CLASH project and knowledge gained from earlier research projects. Since its release, there is additional information (mainly wave overtopping of steep slopes up to vertical structures) and advances made in current practice that are presented in the second edition (EurOtop II, pre-released in December 2016). These manuals provide advice on overtopping predictions for seawalls, embankments, breakwaters and other structures. Although they consider tidal rivers as well as coastal environments, the source of overtopping is from random wave trains associated with wind seas that can cause considerable damage to infrastructure and/or human fatalities in extreme weather conditions. Although management of vessel wash has received attention due to the adverse impacts it can have on safety and the environment, they do not contain formulae to predict overtopping due to vessel-generated waves.

This paper presents a method for modifying the EurOtop II formulae that allows for assessment of overtopping events at vertical river walls (with no influence of the foreshore and non-impulsive conditions) for differing characteristics of vessel wash. Also, it presents physical modelling carried out to validate the approach.

1. BACKGROUND

1.1 Wave overtopping

The formulae presented in this paper for estimating mean discharges and maximum volumes of overtopping due to vessel wash are derived from those given in EurOtop II. The methodology adopted is applicable to plain vertical walls with no influence of the foreshore. The formulae selected correspond to the deterministic approach as this provides additional margin of safety when estimating mean

38

discharges and maximum volumes of overtopping per metre run of wall. Chapter 7 in EurOtop II presents guidance for the assessment of overtopping and post overtopping processes at vertical and steep-fronted coastal structures. The key parameters used to characterize overtopping are mean discharge and maximum volume.

Eurotop II provides Equation 1 to calculate the mean overtopping discharges, which is recommended for designs and safety assessments of overtopping at vertical walls with no influence of the foreshore.

𝑞

√𝑔∗𝐻𝑚03

= 0.054 ∗ 𝑒𝑥𝑝 [− (2.12𝑅𝑐

𝐻𝑚0)

1.3

] Equation 1 (Equation 7.2, EurOtop II)

Regarding the maximum overtopping volume, to estimate this value, first the number of overtopping waves needs to be calculated. For non-impulsive conditions Eurotop II recommends using Equation 2 and 3. These formulae are given by Franco et al. (1994).The number of overtopping waves in a wave set is defined by Equation 2, which assumes a Rayleigh distribution (Weibull distribution with a shape factor of 2) due to being based on a Rayleigh distributed set of waves [Franco et al., 1994].

𝑃𝑜𝑣 =𝑁𝑜𝑤

𝑁𝑤= 𝑒𝑥𝑝 {−

1

𝐶2 (𝑅𝑐

𝐻𝑚0)

2

} for h2/(Hm0*Lm-1,0) > 0.23 Equation 2 (Equation 7.29, EurOtop II)

Where C is defined by Equation 3.

𝐶 = {0.91

0.91 − 0.00425 𝛽0.74

𝑓𝑜𝑟 𝛽 = 0°𝑓𝑜𝑟 0° < 𝛽 < 40°

𝑓𝑜𝑟 𝛽 ≥ 40° Equation 3 (Equation 7.30, EurOtop II)

Finally, Equation 4 is recommended for the calculation of the maximum volumes.

𝑉𝑚𝑎𝑥 = 𝑎(𝑙𝑛 𝑁𝑜𝑤)1/𝑏 Equation 4 (Equation 7.27, EurOtop II)

Where a is defined by Equation 5.

𝑎 = (1

𝛤(1+1

𝑏)) (

𝑞𝑇𝑚

𝑃𝑜𝑣) Equation 5 (Equation 5.53, EurOtop II)

The values a and b in Equation 4 are the scale and shape parameters respectively of the Weibull distribution that EurOtop assumes for the distribution of overtopping volumes. Hughes et al. (2012) and Zanuttigh (2013) describe recent improvements in defining wave overtopping processes using data from experiments with irregular waves. Zanuttigh (2013) gives for b the relationship given in Equation 6.

𝑏 = 0.73 + 55 (𝑞

𝑔𝐻𝑚0𝑇𝑚−1,0)

0.8

Equation 6 (Equation 5.54, EurOtop II)

1.2 Vessel Wakes

1.2.1 Characteristics of the Wake Vessels

The wave system generated by a vessel was described by Lord Kelvin in 1887 [PIANC, 2003]. The wave pattern generated is defined by two types of waves (see Figure 1): diverging waves (propagating at an angle, θ, of 90° to 35° from the vessel’s track) and transverse waves (propagating in the same direction as the vessel). The wave pattern shows a wedge form with an apex angle of approximately 19.5°.

39

The heights of vessel-generated waves depend on bow and stern geometry, vessel speed and dimensions of the chanel (depth and width). The direction and period of the vessel wake waves depend on the vessel path, speed and the water depth. Extensive explanations on vessel wakes and how to calculate them is given by references such as Guidelines for Managing Wake Wash from High-Speed Vessels, PIANC (2003).

Figure 1: Wave crest pattern generated at a vessel bow moving over deep water [Sorensen, 1997]

The wave pattern generated by a single point source in shallow water was investigated by Havelock (1908). He stated that the vessel wake waves in depth-limited water are a function of the depth Froude number. According to this study, the classical Kelvin wave pattern is generated at depth Froude numbers under 0.57. When the depth Froude number approximates 1.0, the vessel’s speed is referred to as the critical speed (Figure 2, left, shows this phenomenon). At this point, vessel wakes build perpendicular to the vessel as transverse waves are left behind. The transverse waves disappear at higher depth Froude numbers.

Figure 2 (left) shows the relationship between the measured vessel-generated maximum wave height and the depth-Froude number [Kirkegaard et al., 1998]. The data is not clearly defined around the trend line shown in the plot; however, the maximum wave heights appear at near below 1.0 depth-Froude numbers (at the critical speed). Also, Figure 2 shows that the vessel-generated wave heights are not larger for super-critical speeds than for sub-critical speeds. Therefore, vessel speeds that go beyond the critical speeds will not be a concern when assessing overtopping events.

Figure 2: Maximum wave height of the long-periodic wave versus the depth-Froude number (left). Maximum wave height of the long-periodic waves versus the distance from the ship track (right). The continuous line indicates the

trend [Kirkegaard, Kofoed-Hansen, & Elfrink, 1998].

40

Figure 2 (right) shows the relationship between the vessel-generated wave heights and the perpendicular distance to the vessel’s track [Kirkegaard et al., 1998]. The reduction in wave height seen for longer distances from the vessel track have no influence from the foreshore and are only influenced by the diffraction of the vessel wake waves.

1.2.2 Statistical Description of Vessel-Generated Waves

The statistical distribution of vessel wakes is one of the most relevant aspects to be addressed in estimation of the maximum overtopping volume. As research done so far regarding overtopping is with wind-sea waves, their statistical differences from the vessel waves considered must be taken into account when applying the methodology given in EurOtop II.

Records of vessel-generated waves are not common and there is no extensive literature on the

statistical distributions of this wave type. Didenkulova and Rodin (2013) present an experimental study

of the group structure of vessel-generated waves. The vessels considered in the study are

approximately 200 m long with operational speeds of 30 knots. It is concluded that Weibull probability

distribution function seems to be a fairly good model describing distribution of wave heights within a

single vessel wake. The scale parameter () is identified between 0.15 and 0.16 m, and the shape

parameter (b) is identified to be 1.45, 1.52 and 1.71 for the different vessels (

Figure 3 shows the Weibull distribution curve for one of the cases). Hence, it appears that vessel wakes

could be defined using the same type of probability distribution as for wind waves in EurOtop if sufficient

wave records were available. However, wave records within waterways are very limited and might not

represent all the vessels expected. Therefore, a different approach may be applied to obtain the wave

distribution that still applies the statistical approach used by EurOtop (see 3.2 for further explanation).

Figure 3: Averaged distribution of wave heights within the wave wake of the three studied vessels. Red dashed

line corresponds to the Rayleigh distribution ( = 0.12 m) and the black solid line corresponds to the Weibull

distribution ( = 0.15 m and b = 1.45) [Didenkulova and Rodin, 2013]

2 PROPOSED METHODOLOGICAL APPROACH

This paper presents a method for assessment of overtopping due to vessel wakes, as currently there is not appropriate guidance on this matter. The methodology proposed is shown in the flow chart in Figure 4.

41

Figure 4: Methodology flow chart for the assessment of overtopping from vessel-generated waves

2.1 Vessel Wake Estimation

For application of this methodology, an actual record of vessel wake waves was not used and they are

often not available. Instead, an estimate of maximum wave height together with a Boussinesq-type wave

train [David et al., 2017] was considered for which the spectral significant wave height, Hm0 (assumed

equal to the significant wave height of the wave train) was determined.

The Coastal Engineering Manual (2006) provides values of maximum vessel-generated wave height for

various vessel types travelling at different speeds (see Table 1).

Characterisation of site conditions:

• Type of structure.

• Influence of the foreshore

Characterisation of vessel wakes:

• Frequency and nature of events - Wave train distribution; o Weibull shape (b) and scale (a)

factors, Spectral significant wave height (Hm0).

o Mean wave period (Tm) - Angle of wave attack (obliquity, β )

Modification of EurOtop Formulae (plain

vertical walls - no influence of the foreshore)

Estimation of overtopping events:

• Mean overtopping discharge (q)

• Maximum overtopping volume (Vmax)

Validation of overtopping estimates:

• Physical model tests

Section 3.1

Section 3.2

Section 3.2

Section 3.3

42

Table 1: Vessel-Generated Wave Heights (shown in Table II-7-5, Coastal Engineering Manual 2006)

Besides the above, there are several models available for the prediction of vessel-generated waves. [Sorensen, 1997] summarises some of these and according to his study only three models among the ones described passes the qualitative evaluation (see Figure 5). These are: Gates & Herbich (1997), Permanent International Association of Navigation Congresses (1987), and Weggel & Sorensen (1986). The latter is considered the best among the three as it considers further vessel characteristics. See

Figure for the comparison between these models and field data. Recently, Macfarlane (2014) created a vessel-generated waves prediction model for vessel operations within sheltered waterways. When sufficient data is available, it is recommended using these models as they will provide a more accurate estimate of the wave train heights.

Figure 5: Ofuya (1970) field data compared with predicted results from PIANC Congresses (1987), Gates & Herbich (1997) and Weggel & Sorensen (1986) models. Chart generated by Sorensen (1997).

43

The wave record considered for this study and used in physical modelling is shown in Figure 6. As mentioned previously, this is the typical response of a Boussinesq-type model to a simplified moving pressure disturbance given by David et al. (2017) in a study of the generation and propagation of vessel-generated waves. This wave set consists of 10 waves and it has a Weibull shape parameter 1.3382, which is slightly less but comparable to the ones found by Didenkulova & Rodin (2013).

Figure 6: Vessel-generated wave record input condition for physical model (Source: David et al. (2017))

2.2 Application and Modification of Eurotop II

To account for the obliquity of waves with respect to the vertical wall, Equation 1 in this paper is modified (see Equation 7) in the same way as Equation 7.16 in EurOtop II by including the factor γβ defined by Equation 8. This modification is required because EurOtop II does not include an equation that accounts with the obliquity of waves for the estimation of overtopping at vertical walls with no influence from the foreshore.

𝑞

√𝑔∗𝐻𝑚03

= 0.054 ∗ 𝑒𝑥𝑝 [− (2.12

𝛾𝛽

𝑅𝑐

𝐻𝑚0)

1.3

] Equation 7

𝛾𝛽 = 1 − 0.0033 ∗ |𝛽| for: 0° ≤ β ≤ 80° (short − crested waves)

Equation 8 (Equation 5.29, EurOtop II)

𝛾𝛽 = 0.736 for: |β| > 80°

Due to the significant spatial variability of overtopping discharge along a seawall or breakwater EurOtop II gives a recommendation for the reduction of the incident angle of the waves (this is defined by Equation 9).

for 𝛽 = 15° ; as per 𝛽 = 0° (𝐸𝑞𝑠. 7.7, 7.8)

𝑓𝑜𝑟 𝛽 = 30° ; 𝑎𝑠 𝑝𝑒𝑟 𝛽 = 15° (𝐸𝑞. 7.18) Equation 9 (7.20, EurOtop II)

for 𝛽 = 60° ; as per non − impulsive 𝛽 = 0° (𝐸𝑞𝑠. 7.7, 7.8)

For risk assessment purposes, the maximum overtopping volume is the value taken into acount and it needs to be calculated. For this, it is necessary to obtain first the number of overtopping waves expected. Although the number of waves of the studied wave train is 10, this would be for a single vessel passing. A larger number of vessel passes should be considered to account for potential traffic on a waterway during the period being assessed. A value for Nw=200 was used in this study for reasons explained

44

below although it could be far higher. To account for the distribution of the wave train wave heights, Equation 2 should be modified to Equation 10 by including the Weibull distribution shape parameter of the wave train proposed (i.e. b= 1.3382) instead of the assumed Rayleigh distribution where the shape parameter is 2 (see b in Equation 10).

𝑵𝒐𝒘

𝑵𝒘= 𝒆𝒙𝒑 {−

𝟏

𝑪𝒃 (𝑹𝒄

𝑯𝒎𝟎)

𝒃

} for h2/(Hm0*Lm-1,0) > 0.23 Equation 10 (Equation 7.29, EurOtop II)

When calculating the maximum overtopping volume (using Equation 4 & 5), the Weibull shape parameter of the vessel wave train is used instead of the one defined by Equation 6, which has been defined using data for irregular waves. With number of waves equal to 10, physical modelling results show that this approach is applicable for the estimation of maximum overtopping volumes (see Figure 9). However, limiting Nw to a few waves means that Equation 4 used to determine Vmax does not provide values at lower mean discharges (higher relative freeboards) as the number of overtopping waves Now is less than one. By increasing Nw to 200 or more and retaining b = 1.3382 would generate very much higher Vmax than observed in physical modelling and likely in the field. Therefore, assuming a repetition of the same wave trains and little spread in Vmax for a large number of passing events, the Weibull shape parameter has been set to b = 5 in Equations 4 and 5.

Further to the above, Figure 7 shows how the Weibull distribution curve with b = 1.3382 fits the data for wave train studied, however, the maximum values at the righthand end of the tail extend much further than for b = 5 (area circled in orange) and beyond the stated Hmax = 0.6 m.

Figure 7: Wave train heights distribution compared to Weibull distributions. Wave train heights distribution is shown by the blue shaded area. The red and green lines show Weibull distributions with shape parameters of

1.3382 and 5 respectively.

As the purpose of Equation 4 is to estimate the maximum overtopping volume associated with maximum wave height, it is considered appropiate to adopt a Weibull distribution shape parameter b = 5 when using this formula (together with Equation 5), as it may describe better the extreme values of the extreme overtopping values. Notwithstanding, where the maximum wave height for vessel wash has been obtained from limited data with potential for more extreme values, a lower value for b may be adopted for added conservatism.

No modification of Equation 5 is proposed that derives the Weibull distribution scale factor.

2.3 Physical Modelling Validation

Physical modelling was undertaken to investigate the difference in overtopping volumes between standard vessel-generated waves and irregular wind-sea waves with a view to validating use of EurOtop overtopping formulae modified for vessel waves. The tests were carried out in a wave basin, which measures approximately 32 m by 20 m. During the physical modelling both irregular (JONSWAP

45

spectrum) and representative vessel wake waves (given by David et al. (2017) and shown in Figure 6) were used. The conditions tested in the experiments are shown in Table 2 and 3.

Wave condition Maximum wave height (m) Wave period (s)

Vessel 1 0.6 3.3

Table 2: Overview vessel wave conditions used in the physical model

Wave condition Significant wave height (m)

Wave period (s) JONSWAP gamma

SeaStateA 0.6 3.3 3.3

Table 3: Overview wind-generated wave conditions used in the physical model

It was found that mean overtopping discharges by irregular sea waves and vessel waves impacting on a plain vertical wall follow the same trend for different ratios of freeboard with wave height (see Figure 8). This supports use the EurOtop II formulae defining overtopping of irregular wind waves to estimate overtopping due to vessel wakes albeit with selection of appropriate parameters to describe the wave train. The mean discharges measured during these experiments are considerably larger than the ones predicted by Equation 7 and CLASH data. This may be due to the limited number of experiments carried out whilst a high variability of mean discharges is indicated from previous studies of irregular waves. Additional physical modelling could be undertaken to obtained more data, but the results are considered sufficient to support the premise that both irregular sea waves and vessel wash have a similar relationship for mean overtopping discharges with relative freeboard.

Figure 7: CLASH database for vertical walls without influence from the foreshore [EurOtop, 2016]. The

overtopping tests combined trend line (vessel waves and irregular waves) is included. Eq. 1 & 2 mentioned in the

figure are found in EurOtop II.

Figure 9 shows that the estimated maximum overtopping discharges obtained by modifying the parametres suggested in EurOtop (modifications shown in Section 3.2) correlate well with the physical modelling observations for Nw= 10. It is also shown that provided formulae estimate significantly higher maximum overtopping volumes if 200 waves were considered. As explained in Section 3.2 this is due to the variability of extreme values when lower Weibull shape parameters are considered (b= 1.3382 in this case). Figure 10 shows that a higher Weibull shape parameter (i.e. b= 5) considering 200 waves

46

will give comparable results to the physical modelling measurements (for single wave train consisting of 10 waves). This should be applied when less variability of the extreme values is expected.

Figure 9: Relationship mean discharge against maximum overtopping volume - Wave Obliquity 35 degrees; Nw = 10 & 200; b=1.3382 in Equation 10 and Equation 4.

Figure 10: Relationship mean discharge against maximum overtopping volume – Wave Obliquity 35 degrees; Nw = 200; b=1.3382 in Equation 10; b=5 in Equation 4.

Figure 9 shows that EurOtop formulae underestimate maximum overtopping volumes for lower values when considering a small number of waves (see red dot for a maximum overtopping volume of 5 litres). This happens when the number of overtopping waves is close to 1. By considering a larger number of waves the values underestimation is avoided (see Figure 9 and 10 for Nw= 200). Also, the maximum overtopping volumes measured in the physical modelling and those estimated using the formulae shown deviate for smaller volumes (see physical modelling results for volumes below 10 litres) when considering 200 waves (physical modelling measurements drop rapidly and lose linearity). A possible explanation is that the measurement equipment considerably loses accuracy for low volumes, as there

47

is no evidence in literature for a drop on the trend’s linearity in logarithmic scale. In any case, these lower maximum overtopping volumes are not significant for the assessment of overtopping discharges.

3 SUMMARY

This paper presents modified EurOtop formulae to estimate overtopping due to vessel-generated waves at vertical walls alongside navigable rivers and waterways. To the authors’ knowledge, such an approach has not been presented previously with research of overtopping being largely focused on wind-sea waves. The method for estimation of overtopping from vessel waves has been validated by physical modelling for the specific case studied (i.e. non-impulsive waves overtopping a plain vertical wall without influence of the foreshore). It partly fills a gap in coverage and could be applied to other cases of vessel waves overtopping with different types of structures subject to further physical tests to verify the results.

4 SYMBOLS

q Mean overtopping discharge

g Gravity

Hmo Spectral sig. wave height

Rc Freeboard

Now Number of overtopping waves

Nw Number of waves

C C parameter

L Wavelength

Wave obliquity

Vmax Max overtopping volume

b Weibull shape parameter

a Weibull scale parameter

Ganma function

Pov Probability of overtopping

T Wave period

Influence factor

H Wave height

5 REFERENCES

Coastal Engineering Manual (2006): "Washington, D.C.: U.S. Army Corps of Engineers".

David, C. G., Roeber, V., Goseberg, N. and Schlurmann, T. (2017): "Generation and Propagation of

Ship-Borne Waves-Solutions from a Boussinesq-Type Model", Coastal Engineering, 170-187.

Didenkulova, I. and Rodin, A. (2013): "A Typical Wave Wake from High-Speed Vessels: Its Group

Structure and Run-Up", Nonlinear Processes in Geophysics, 179-188.

EurOtop II (2016): "Manual on Wave Overtopping of Sea Defences and Related Structures. An

Overtopping Manual Largely Based on European Research, But for Worldwide Application. European

Overtopping Manual".

Franco, L., de Gerloni, M. and van der Meer, J. W. (1994): "Wave Overtopping on Vertical and

Composite Breakwaters", Proc. 24th International Conference on Coastal Engineering, (pp. 1030-1044).

Kobecuantar.

Gates, E. T. and Herbich, J. B. (1997): "Mathematical Model to Predict the Behavior of Deep-Draft

Vessels in Restricted Waterways", Texas A&M University, College Station, TX.

48

Havelock, T.H. (1908): "The Propagation of Groups of Waves in Dispersive Media, with Application to

Waves on Water Produced by a Travelling Disturbance", Proc. R. Soc. Lond. A, 81(549), pp.398-430.

Hughes, S., Thornton, C., Van der Meer, J. W. and Scholl, B. (2012): "Improvements in describing wave

overtopping processes". Proc. ICCE, ASCE.

Kirkegaard, J., Kofoed-Hansen, H. and Elfrink, B. (1998): "Wake Wash of High-Speed Craft in Coastal

Areas", Coastal Engineering.

Macfarlane, G.J., Bose, N. and Duffy, J.T. (2014): "Wave Wake: Focus on Vessel Operations within

Sheltered Waterways", Lauceston, Tasmania, Australia: Australian Maritime College, University of

Tasmania.

Ofuya, A.O. (1970): "Shore Erosion - Ship & Wind Waves: St Clair, Detroit & St. Lawrence Rivers",

Report 21. Canada: Marine Engineering Division, Department of Public Works.

Permanent International Association of Navigation Congresses (PIANC) (1987): "Guidelines for the

Design and Construction of Flexible Revetments Incorporating Geotextiles for Inland Waterways",

Brussels: Working Group 4 of the Permanent Technical Committee.

PIANC (2003): "Guidelines For Managing Wake Wash From High-Speed Vessels", Brussels: PIANC

General.

Pullen, T., Allsop, N. W. H., Bruce, T., Kortenhaus, A., Schüttrumpf, H. and Van der Meer, J. W. (2007):

"EurOtop, European Overtopping Manual-Wave Overtopping of Sea Defences and Related Structures:

Assessment Manual", also published as Special Volume of Die Küste.

Sorensen, R.M. (1997): "Prediction of Vessel-Generated Waves with Reference to Vessels Common to

the Upper Mississipi River System", Bethlehem: Department of Civil and Environmental Engineering,

Lehigh University.

Weggel, H.J. and Sorensen, R.M. (1986): "Ship Waves Prediction for Port and Channel Design",

Proceedings of the Ports '86 (pp. 797-814), New York: American Society of Civil Engineers.

Zanuttigh, B., Van der Meer, J.W., Bruce, T. and Hughes, S. (2013): "Statistical Characterisation of

Extreme Overtopping Volumes", Proc. ICE, Coasts, Marine Structures and Breakwaters 2013,

Edinburgh, UK.

49

SUMMARY

Riverside walkways and promenades are very common in towns and cities with navigable waterways.

Some of these locations are exposed to overtopping of vessel-generated waves. The widely used

guidance to assess overtopping is EurOtop, which is applicable to random wave trains associated with

wind seas. Currently there is not a methodology for the assessment of overtopping due to vessel wash.

This paper presents a procedure for estimation of overtopping due to vessel wake waves at vertical

walls (with no influence of the foreshore and non-impulsive conditions) by modifying the equations given

in EurOtop II (pre-released 2016). The alterations made address statistical differences between wind-

generated and vessel-generated waves, by changing the Weibull scale and shape parameters that

describe the distribution of overtopping events. Physical modelling was carried out to validate the

approach. The modelling results indicate that mean overtopping discharge from both irregular and ship

waves are comparable for the case studied. It was concluded that the modified EurOtop II formulae are

appropriate for estimating the mean discharge and maximum volume of overtopping due to wake waves.

RESUME

Les promenades et les promenades au bord de l'eau sont très fréquentes dans les villes où il y a des voies navigables. Certains de ces endroits sont exposés au débordement des vagues générées par les bateaux. Le guide largement utilisé pour évaluer le débordement est EurOtop, qui s'applique aux trains de vagues aléatoires associés aux mers de vent. Actuellement, il n'existe pas de méthodologie pour l'évaluation du débordement dû au lavage des navires. Ce document présente une procédure d'estimation du débordement dû aux vagues de sillage des navires sur les parois verticales (sans influence de l'estran et des conditions non impulsives) en modifiant les équations données dans EurOtop II (pré-publié en 2016). Les modifications apportées portent sur les différences statistiques entre les vagues générées par le vent et celles générées par les navires, en changeant l'échelle de Weibull et les paramètres de forme qui décrivent la distribution des événements de débordement. Une modélisation physique a été réalisée pour valider l'approche. Les résultats de la modélisation indiquent que le déversement moyen des vagues irrégulières et des vagues de navire est comparable pour le cas étudié. Il a été conclu que les formules EurOtop II modifiées sont appropriées pour estimer la décharge moyenne et le volume maximum de débordement dû aux vagues de sillage.

ZUSAMMENFASSUNG

Spazierwege und Promenaden entlang von Flüssen sind in Gemeinden und Städten, die an Wasserstraßen liegen, weit verbreitet. Manche dieser Wege sind einem von Schiffen verursachten Wellenüberlauf ausgesetzt. Der am häufigsten verwendete Leitfaden, um einen Wellenüberlauf abzuschätzen, ist EurOtop, der auf beliebige, im Zusammenhang mit der Windsee stehende, Wellenzüge angewendet werden kann. Zurzeit gibt es keine Methode für die Bewertung eines durch Schiffswellen verursachten Wellenüberlaufs. Dieser Artikel präsentiert ein Verfahren zur Abschätzung des Wellenüberlaufs durch das Kielwasser eines Schiffes an senkrechten Wänden (ohne Einfluss des Vorlandes und unter impulsfreien Bedingungen), indem die Gleichungen des EurOtop II (vor-veröffentlicht 2016) modifiziert wurden. Die vorgenommenen Änderungen berücksichtigen statistische Unterschiede zwischen wind- und schiffserzeugten Wellen, indem die Weibull Skalen- und Formparameter, die die Verteilung des Wellenüberlaufs beschreiben, verändert wurden. Physikalische Modelluntersuchungen wurden durchgeführt, um den Ansatz zu validieren. Die Modellergebnisse zeigen, dass der mittlere Abfluss aus dem Überlauf sowohl durch irreguläre Wellen als auch Schiffswellen für den betrachteten Fall vergleichbar ist. Es wurde gefolgert, dass die modifizierten EurOtop II Formeln geeignet sind, um den durchschnittlichen Abfluss und das Maximalvolumen des Wellenüberlaufs durch Kielwasser abzuschätzen.

50

RESUMEN

En las ciudades en las que existen vías navegables es habitual que existan zonas de paseo contiguas a sus márgenes. Algunas de estas zonas están expuestas a la aparición de fenómenos de rebase producidos por el oleaje que generan los propios buques durante su navegación. La forma habitual de evaluar el rebase es a través de “EurOtop”, metodología que es aplicable a condiciones de oleaje aleatorio generado por la acción del viento sobre la superficie de agua. Actualmente no existe una metodología para evaluación del rebaje por oleaje generado por buques. Este artículo muestra un procedimiento para estimar este tipo de situaciones para el caso de paredes verticales en las que no existe influencia de la costa y no se producen condiciones impulsivas, modificando las fórmulas que aparecen en “EurOtop II” (lanzado en 2016). Las modificaciones propuestas se deben a las diferencias de caracterización existentes entre los oleajes generados por viento y los generados por buques, modificando la distribución Weibull y los parámetros de forma que gobiernan la distribución de los eventos de rebase. Para validar esta aproximación se han desarrollado ensayos en modelo físico. El modelo desarrollado indica que los caudales generados en las situaciones de oleaje irregular y de oleaje generado por buques resultan ser comparables. Se puede concluir que las fórmulas modificadas de “EurOtop II” son adecuadas para valorar caudales y volúmenes máximos en situaciones de rebase producidas por olas procedentes de buques.